Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Convergence test - wave optics interface

Krishna Koundinya Upadhyayula

Please login with a confirmed email address before reporting spam

considering the scattering problem off of a si nanosphere: a si nanosphere radius 250nm and a pml radius of 2300 nm. Incident radiation with lambda = 1000nm propagating in z direction has an electric field component in the x direction E0*exp(ikz). and E0 = 1 V/m.

why does the power outflow time averaged plotted against the thickness of pml layer stay constant for increasing pml thickness up until about 1.5 times the incident wavelength and decrease sharply later? (or alternatively, power outflow time averaged vs the radius of the airdomain outside the nano-particle, excluding the pml layer - increase with increase in physical domain radius or the surface area of the air -domain and then reaches a constant value at about a radius 1.5 times the incident wavelength?)

Considering that the time averaged power outflow is the integral of the Poynting vector over the surface area of the physical domain (the pml absorbs the radiation entering it, making integral over the entire model domain redundant), this should mean that the scattering energy increases with the surface area of the air domain outside the nanoparticle- (excluding the pml layer). what then, is expected to happen as the physical domain surface area increases to infinity?? ![https://imgur.com/oHYxy4S]


1 Reply Last Post Feb 20, 2020, 9:50 a.m. EST
Lars Dammann COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 4 years ago Feb 20, 2020, 9:50 a.m. EST

Hi, I'm not really sure what you have integrated there (inside or outside of the PML?), but the value looks to be on the order of 1e-17, so really small. I think what you are seeing could just be numerical noise, so different approximations of zero. In that case there is not much to be learned from those numbers. Is there any particular reason you are wondering about this? Do your solutions actually look different, when you change the PML thickness? Generally, the PML thickness mildly influences the effectiveness of the absorption. I usually use one wavelength geometrical size and 5 mesh elements using a structured mesh as a rule of thumb. I could probably give a more detailed answer, if you could upload your model file here. Thanks and best regards, Lars

Hi, I'm not really sure what you have integrated there (inside or outside of the PML?), but the value looks to be on the order of 1e-17, so really small. I think what you are seeing could just be numerical noise, so different approximations of zero. In that case there is not much to be learned from those numbers. Is there any particular reason you are wondering about this? Do your solutions actually look different, when you change the PML thickness? Generally, the PML thickness mildly influences the effectiveness of the absorption. I usually use one wavelength geometrical size and 5 mesh elements using a structured mesh as a rule of thumb. I could probably give a more detailed answer, if you could upload your model file here. Thanks and best regards, Lars

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.